Sequencing of the genome of Anopheles gambiae s.s., one of the primary mosquito vector of malaria in Africa, has been touted as ...a breakthrough in public health by providing a means for mapping, selecting and exploiting genes of interest. To date the major focus of such efforts has been on exploring the genetic basis of susceptibility to malaria parasites (Plasmodium sp.) in model systems in the laboratory. These have often not translated well to what is occurring in nature. Indeed, it has become clear that examining vector phenotypes in an ecological context, as they occur in nature, is critical for producing results relevant to malaria epidemiology in real transmission settings. In this application we propose a research program that integrates vector population genomics, ecology and vector behavior with the goal of understanding the determinants of two mosquito behavioral phenotypes crucial to the transmission and control of malaria: (1) host-preference and (2) adult resting behavior. Our approach builds upon a sizeable base of preliminary work in the laboratory which has identified an extensive panel of An. arabiensis SNP markers, and preliminary field work in Tanzania that has identified a range of appropriate sites where sampling methods have been piloted and the behavior of An. arabiensis is known to vary. An. arabiensis mosquitoes will be intensively collected from four villages in the Kilombero Valley of Tanzania during the wet and dry seasons to determine the association between their feeding and resting phenotype and environmental factors that vary temporally and spatially (Aim #1). DNA will be extracted from individual samples and multi-locus SNP genotypes determined from each individual. Genotypes will be organized by phenotype (exophilic vs. endophilic and human fed vs. animal fed) and analyzed to determine SNP allele associations with each phenotype after correcting for population structure and genetic relatedness using the efficient mixed-model association (EMMA) method (Aim #2). Knowledge of the genetic basis of these behavioral changes will be vital for prediction of both possible downstream evolutionary responses to current vector control strategies, and also for the development of novel control strategies that improve the application of currently available vector control methods and/or that are based on vector genetic manipulation. We have enlisted experts in the fields of association mapping (Dr. Eleazar Eskin) and spatial analysis of ecological data (Dr. Daniel Haydon) who will serve as consultants to ensure that the most contemporary and best methods are used for the analysis of data.